Tumourous stem mustard, an important cash crop and the raw material for Fuling pickles, is a world-famous vegetable crop. Presently, many varieties of tumourous stem mustard have been bred, but the regulatory pathway and molecular mechanism of mustard tumourous stem development are unclear, and molecular biological research on tumourous stem mustard is rare.
By transcriptome sequence analysis, we obtained 54,577,780 reads corresponding to about 5 Gb of raw sequence data. The predicted 146,265 unigenes were subjected to BLAST annotation, and 72% of the unigenes returned a significant BLAST result. As expected, most unigenes shared the highest sequence similarity with crucifers (A. thaliana, A. lyrata subsp. lyrata, and Brassica). The number of genes similar to Brassica genes was lower than the number of similar genes between A. thaliana and A. lyrata subsp. lyrata (Figure 2), possibly due to the characteristics of the B. juncea transcriptome in the NCBI protein database. Arabidopsis thaliana is an important model plant with a clear genetic background that is very useful for researching gene functions in tumourous stem mustard. Our transcriptome analysis is the first high-throughput sequencing of tumourous stem mustard and will serve as a basis for other studies.
To investigate the regulatory pathway and molecular mechanism of tumour swelling, we created five DGE libraries from plants at different developmental stages and samples from a non-swollen mutant to analyse the gene expression patterns at various developmental stages. The quality of the DGE libraries was further confirmed by qRT-PCR analysis. Because the tumours continue to swell for a long period after initiation, different swelling stages were selected for each experimental group, and non-swollen stages of Yong'an and the non-swollen mutant strain Dayejie were used for comparison. The DGE profiles of the swelling stages were compared with the controls. Different genes may be involved in tumourous stem formation, and the common genes in the six comparison groups reduced the number of differentially expressed genes that might be related to tumourous stem development.
Compared with the GO annotation results of the DGE screening of genes in the transcriptome data, we found no genes distributed in the three molecular function subcategories (enzyme regulator activity, structural molecular activity, or translation regulator activity) in the DGE group, indicating that these three subcategories are not related to tumour swelling. A comparison of the results of Additional file 3: Table S3 and Table S6 showed that the pathway order number of 'Plant hormone signal transduction' was about 10 (Additional file 6: Table S6), suggesting that these genes were screened out and that the pathway 'plant hormone signal transduction' is related to tumour swelling. Mapping the DGE data back to the transcriptome database revealed that about 30% of the reads were mapped and that > 60% remained transcribed sequences. Although large amounts of data were obtained by transcriptome sequencing, the reference sequences may still be insufficient and may have caused the lower mapped ratio, which could be resolved by increasing the sequencing depth and enhancing the accuracy of the assembly.
Although 1,042 differentially expressed genes were discovered using the above method, the key genes related to tumourous stem formation need to be analysed further. The genes with the greatest changes in expression were selected for further study; a log 2 ratio value > 10 was used as a threshold to select thirteen genes for further analysis. Seven of the genes have unknown functions and six genes have a functional annotation based on sequence similarity. Of the six annotated genes, four gene functions require further clarification, and two genes, phytochrome kinase substrate 1 (PKS1) (Unigene142399_num2_yongan) and ABR1 (ABA REPRESSOR1) (Unigene140028_num2_yongan), have functions whose details are relatively well known. Girdhar et al.  determined that ABR1 functions in the negative regulation of ABA responses during seed germination and plays a role in the ABA signalling pathway in Arabidopsis. In this study, the expression level of ABR1 was significantly higher in the inflation stage of the stem tumour. This suggests that ABA signalling is related to stem inflation. As Additional file 6: Table S6 shows, four genes may be involved in the 'plant hormone signal transduction [ko04075]' pathway, suggesting that plant hormones are related to stem swelling; however, ABR1 was not one of the four genes annotated in the 'plant hormone signal transduction' pathway in this study. During the inflation stage of stem tumours, plant hormones play key roles in the number of cells in the stem, which increase rapidly accompanied by cell splintering. PKS1 expression was upregulated during the tumour inflation stage. To our knowledge, PKS1 is phosphorylated in a phytochrome-dependent manner and negatively regulates phytochrome light signalling in Arabidopsis . Other reports have shown that PKS1 regulates root phototropism and gravitropism and leaf flattening and positioning, and that it affects the state of phytochrome A in etiolated Arabidopsis seedlings [16–20]. Chen et al. [21, 22] found that illumination and temperature were the main factors affecting the formation of stem tumours, but they did not determine whether PKS1 plays key roles in illumination. Is PKS1 related to stem inflation in tumourous stem mustard? Whether or not the overexpression or knockout/down of ABR1 and PKS1 alters the trend of inflation requires further research.